Crown of thorns starfish

Description

The body form of the crown-of-thorns starfish is fundamentally the same as that of a typical starfish, with a central disk and radiating arms. Its special traits, however, include being disc-shaped, multiple-armed, flexible, prehensile, and heavily spined, and having a large ratio of stomach surface to body mass. Its prehensile ability arises from the two rows of numerous tube feet that extend to the tip of each arm. In being multiple-armed, it has lost the five-fold symmetry (pentamerism) typical of starfish, although it begins with this symmetry in its life cycle.

Adult crown-of-thorns starfish normally range in size from 25 to 35 cm (9.8 to 13.8 in). They have up to 21 arms. Although the body of the crown of thorns has a stiff appearance, it is able to bend and twist to fit around the contours of the corals on which it feeds. The underside of each arm has a series of closely fitting plates which form a groove and extend in rows to the mouth. They are usually of subdued colours, pale brown to grey-green, but they may be garish with bright warning colours in some parts of their wide range.

The long, sharp spines on the sides of the starfish's arms and upper (aboral) surface resemble thorns and create a crown-like shape, giving the creature its name. The spines are stiff and very sharp, and readily pierce through soft surfaces. Despite the battery of sharp spines on the aboral surface and blunt spines on the oral surface, the crown-of-thorns starfish's general body surface is membranous and soft. When the starfish is removed from the water, the body surface ruptures and the body fluid leaks out, so the body collapses and flattens. The spines bend over and flatten, as well. They recover their shape when reimmersed, if they are still alive.

Family

The family Acanthasteridae is monogeneric; its position within the Asteroides is unsettled. It is generally recognised as a distinctly isolated taxon. Recently, Blake concluded from comparative morphology studies of Acanthaster planci that it has strong similarities with various members of the Oreasteridae. He transferred Acanthasteridae from the Spinulosida to the Valvatida and assigned it a position near to the Oreasteridae, from which it appears to be derived. He attributed Acanthaster morphology as possibly evolving in association with its locomotion over irregular coral surfaces in higher energy environments. There is a complication, however, in that Acanthaster is not a monospecific genus and any consideration of the genus must also take into account another species, Acanthaster brevispinus, which lives in a completely different environment. A. brevispinus lives on soft substrates, perhaps buried in the substrate at times like other soft substrate-inhabiting starfish, at moderate depths where presumably the surface is regular and there is little wave action.

Genus and species

Acanthaster planci has a long history in the scientific literature with great confusion in the generic and species names from the outset, with a long list of complex synonyms. As a very distinctive starfish, it is not surprising that it was first described in 1705. Rhumphius used the name Stella marina quindecium radiotorum. Later, Linnaeus described it as Asterias planci, based on an illustration by Plancus and Gualtieri (1743), when he introduced his system of binomial nomenclature. No type specimens are known; the specimen described by Plancus and Gualtieri (1743) is no longer extant.

Subsequent generic names used for the crown-of-thorns starfish included Stellonia, Echinaster and Echinites, before settling on Acanthaster (Gervais 1841). Species names included echintes, solaris, mauritensis, ellisii, and ellisii pseudoplanci (with subspecies). Most of these names arose from confusion in the historical literature, but Acanthaster ellisii came to be used for the distinctive starfish in the eastern Pacific Gulf of California.

The eastern Pacific Acanthaster is very distinctive (see image to the right) with its rather 'plump' body, large disk to total diameter ratio and short, blunt spines. It gives the impression of living in a habitat where having sharp defenses against predators has little value, although it lives on coral reef and feeds on coral.

Genetic studies

Nishida and Lucas examined variation at 14 allozyme loci of ten population samples of A. planci using starch-gel electrophoresis. The samples were from localities across the Pacific: Ryukyu archipelago (four locations), Micronesia (two locations), and samples from one location each of the Great Barrier Reef, Fiji, Hawaii, and Gulf of California. A sample of 10 specimens of A. brevispinus from the Great Barrier Reef region was included for comparison. There was considerable genetic differentiation between the A. brevispinus and A. planci populations (D= 0.20 +/- 0.02)(D is genetic distance). The genetic differences between geographic populations of A. planci were, however, small (D = 0.03 +/- 0.00; Fsr = 0.07 + 0.02) (Fsr is standardized genetic variance for each polymorphic locus) despite the great distances separating them. A positive correlation was observed between degree of genetic differentiation and geographic distance, suggesting the genetic homogeneity among A. planci populations is due to gene flow by planktonic larval dispersion. The distance effect on genetic differentiation most probably reflects decreasing levels of successful larval dispersal over long distances. In view of the level of macrogeographic homogeneity, significant allele frequency differences were observed between adjacent populations separated by approximately 10 km. The Hawaiian population was most differentiated from other populations. Treating the morphologically distinctive, eastern Pacific Acanthaster as a separate species, A. ellisii, is not supported by these data. The lack of unique alleles in the central (Hawaii) and eastern Pacific (Gulf of California) populations suggests they were derived from those in the western Pacific.

Further details of the genetic relationship between A. planci and Acanthaster brevispinus are presented in the entry for the latter species. These are clearly sibling species and it is suggested that A. planci, the specialised coral-feeding species arose from A. brevispinus, the less specialised soft-bottom inhabitant.

In a very comprehensive geographic study, Benzie examined allozyme loci variation in 20 populations of A. planci, throughout the Pacific and Indian Oceans. The most striking result was a very marked discontinuity between the Indian and Pacific Ocean populations. Those, however, off northern Western Australia had a strong Pacific affinity. With the exception of the very strong connection of southern Japanese populations to the Great Barrier Reef populations, the patterns of variation within regions were consistent with isolation by distance. Again, the pattern of decreasing levels of successful larval dispersal over long distances is apparent. Benzie suggests that the divergence between Indian Ocean and Pacific Ocean populations began at least 1.6 million years ago and is likely to reflect responses to changes in climate and sea level.

A more recent comprehensive geographic study of A. planci by Vogler et al., using DNA analyses (one mitochondrial gene), suggests it is actually a species complex consisting of four species or clades. The four cryptic species/clades are defined geographically: northern Indian Ocean, southern Indian Ocean, Red Sea and Pacific Ocean. These molecular data suggest the species/clades diverged 1.95 and 3.65 million years ago. (The divergence of A. planci and A. brevispinus is not included in this time-scale.) The authors suggest the differences between the four putative species in behaviour, diet or habitat may be important for the design of appropriate reef conservation strategies.

There are, however, problems with this proposal of cryptic speciation (cryptic species). The basis of these data from one mitochondrial gene. mtDNA data are, however, only one source of information about the status of taxa and the use of one mtDNA gene as a sole criterion for species identification is disputed. The allozyme data should also be taken into account. Three localities that were sampled by Vogler et al. are of particular interest: Palau Sebibu, UEA and Oman were found to have two clades/sibling species in sympatry. These are important to investigate the nature of the co-existence and barriers to introgression of genetic material. A. planci as a taxon is a generalist, being amongst the most ubiquitous of large coral predators on coral reefs, feeding on virtually all hard coral species, reproducing during summer without a pattern of spawning and often participating in mass multi-species spawnings, and releasing vast amounts of gametes that trigger spawning in other individuals. It is very difficult to conceive of two species/clades of A. planci in sympatry without habitat competition and introgression of genetic material, especially the latter.

Starfish are characterised by having saponins known as asterosaponins in their tissues. They contain a mix of these saponins, and at least 15 chemical studies have been conducted seeking to characterise these saponins. The saponins have detergent-like properties, and keeping starfish in limited water volumes with aeration results in large amounts of foam at the surface.

A. planci has no mechanism for injecting the toxin, but as the spines perforate tissue of a predator or unwary person, tissue containing the saponins is lost into the wound. In humans, this immediately causes a sharp, stinging pain that can last for several hours, persistent bleeding due to the haemolytic effect of saponins, and nausea and tissue swelling that may persist for a week or more. The spines, which are brittle, may also break off and become embedded in the tissue where they must be removed surgically.

Saponins seem to occur throughout the lifecycle of the crown-of-thorns starfish. The saponins in the eggs are similar to those in the adult tissues, and presumably these carry over to the larvae. The mouthing behaviour of predators of juvenile starfish with rejection suggests the juveniles contain saponins.

The adult crown-of-thorns is a carnivorous predator that usually preys on reef coral polyps. It climbs onto a section of living coral colony using the large number of tube feet on its oral surface and flexible body. It fits closely to the surface of the coral, even the complex surfaces of branching corals. It then extrudes its stomach out through its mouth over the surface to virtually its own diameter. The stomach surface secretes digestive enzymes that allow the starfish to absorb nutrients from the liquefied coral tissue. This leaves a white scar of coral skeleton which is rapidly infested with filamentous algae. An individual starfish can consume up to 6 square metres (65 sq ft) of living coral reef per year. In a study of feeding rates on two coral reefs in the central Great Barrier Reef region, large starfish (40 cm and greater diameter) killed about 61 cm²/day in winter and 357–478 cm²/day in summer. Smaller starfish, 20–39 cm, killed 155 and 234 cm²/day in the equivalent seasons. The area killed by the large starfish is equivalent to about 10 m² (110 sq ft) from these observations. Differences in feeding and locomotion rates between summer and winter reflect the fact that the crown-of-thorns, like all marine invertebrates, is a poikilotherm whose body temperature and metabolic rate are directly affected by the temperature of the surrounding water. In tropical coral reefs, crown-of-thorns specimens reach mean locomotion rates of 35 cm/min, which explains how outbreaks can damage large reef areas in relatively short periods.

The starfish show preferences between the hard corals on which they feed. They tend to feed on branching corals and table-like corals, such as Acropora species, rather than on more rounded corals with less exposed surface area, such as Porites species. Avoidance of Porites and some other corals may also be due to resident bivalve molluscs and polychaete worms in the surface of the coral which discourage the starfish. Similarly, some symbionts, such as small crabs, living within the complex structures of branching corals, may ward off the starfish as it seeks to spread its stomach over the coral surface.

In reef areas of low densities of hard coral, reflecting the nature of the reef community or due to feeding by high density crown-of-thorns, the starfish may be found feeding on soft corals (Alcyonacea).

The starfish are cryptic in behavior during their first two years, emerging at night to feed. They usually remain so as adults when solitary. The only evidence of a hidden individual may be white feeding scars on adjacent coral. However, their behavior changes under two circumstances:

Predators

The elongated sharp spines covering nearly the entire upper surface of the crown-of-thorns serve as a mechanical defense against large predators. It also has a chemical defense. Saponins presumably serve as an irritant when the spines pierce a predator, in the same way as they do when they pierce the skin of humans. Saponins have an unpleasant taste. A study to test the predation rate on juvenile Acanthaster by appropriate fish species found that the starfish were often mouthed, tasted, and rejected. These defenses tend to make it an unattractive target for coral community predators. In spite of this, however, Acanthaster populations are typically composed of a proportion of individuals with regenerating arms.

A variety of about 11 species have been reported to prey occasionally on uninjured and healthy adults of A. planci. All of these are generalist feeders and none of these, however, seems to specifically prefer the starfish as a food source. This number, however, is probably lower, as some of these presumed predators have not been witnessed reliably in the field. Some of those witnessed are:

Gonads increase in size as the animals become sexually mature and at maturity fill the arms and extend into the disk region. The ripe ovaries and testes are readily distinguished with the former being more yellow and having larger lobes. In section they are very different with the ovaries densely filled with nutrient-packed ova (see ovum and photograph) and the testes densely filled with sperm, which consist of little more than a nucleus and flagellum. Fecundity in female Crown-of-thorns starfish is related to size with large starfish committing proportionally more energy into ova production such that a:

In coral reefs in the Philippines, female specimens were found with a gonadosomatic index (ratio of gonad mass to body mass) as high as 22%, which underlines the high fecundity of this starfish. Babcock et al. (1993) monitored changes in fecundity and fertility (fertilisation rate) over the spawning season of the Crown-of-thorns starfish on Davies Reef, central Great Barrier Reef, from 1990 to 1992. The starfish were observed to spawn (photograph) from December to January (early to mid-summer) in this region with most observations being in January. However, both gonadosomatic index and fertility peaked early and declined to low levels by late January, indicating that most successful reproductive events took place early in the spawning season. In northern hemisphere coral reefs however, crown-of-thorns populations reproduce in April and May, and were also observed spawning in the Gulf of Thailand in September. High rates of egg fertilisation may be achieved through the behaviour of proximate and synchronised spawning (see above in Behaviour).

Embryonic development begins about 1.5 hours after fertilisation, with the early cell divisions (cleavage) (photograph). By 8–9 hours it has reached the 64-cell stage.

By Day 1 the embryo has hatched as a ciliated gastrula stage (photograph). By Day 2 the gut is complete and the larva is now known as a bipinnaria (photographs). It has ciliated bands along the body and uses these to swim and filter feed on microscopic particles, particularly unicellular green flagellates (phytoplankton). The SEM photograph is a scanning electron micrograph, which clearly shows the complex ciliated bands of the bipinnaria larva. By Day 5 it is an early brachiolaria larva. The arms of the bipinnaria have further elongated, there are two stump-like projections in the anterior (not evident in the photograph) and structures are developing within the posterior of the larva. In the late brachiolaria larva (Day 11)(photograph) the larval arms are elongate and there are three distinctive arms at the anterior with small structures on their inner surfaces (photographs). To this stage the larva has been virtually transparent, but the posterior section is now opaque with the initial development of a starfish. The late brachiolaria is 1-1.5 mm. It tends to sink to the bottom and test the substrate with its brachiolar arms, including flexing the anterior body to orient the brachiolar arms against the substrate.

This description and assessment of optimum rate of development is based on early studies in the laboratory under attempted optimum conditions. However, not unexpectedly, there are large differences in growth rate and survival under various environmental conditions (see Causes of population outbreaks).

The late brachiolaria search substrates with their arms and, when offered a choice of substrates, tend to settle on coralline algae, which they will subsequently feed on. In the classic pattern for echinoderms, the bilaterally symmetrical larva is replaced by a pentamerously symmetrical stage at metamorphosis, with the latter's body axis bearing no relationship to that of the larva. Thus the newly metamorphosed starfish are five-armed and are 0.4–1 mm diameter. (Note the size of the tube feet relative to the size of the animal.) They feed on the thin coating layers of hard encrusting algae (coralline algae) on the undersides of dead coral rubble and other concealed surfaces. They extend their stomach over the surface of the encrusting algae and digest the tissue, as in the feeding by larger crown-of-thorns starfish on hard corals. The living tissue of the encrusting algae is approximately pink to dark red and feeding by these early juveniles results in white scars on the surface of the algae (photograph). During the next months, the juveniles grow and add arms and associated madreporites in the pattern described by Yamaguchi until the adult numbers is attained 5–7 months after metamorphosis. Two hard corals with small polyps, Pocillopora damicornis and Acropora acunimata, were included in the aquaria with the encrusting algae and at about the time the juvenile starfish achieved their full number of arms they began feeding on the corals.

Juvenile A. planci that had reached the stage of feeding on coral were then reared for some years in the same large closed-circuit seawater system that was used for the early juveniles. They were moved to larger tanks and kept supplied with coral so that food was not a limiting factor on growth rate. The growth curves of size versus age were sigmoidal, as seen in majority marine invertebrates. There was an initial period of relatively slow growth while the starfish were feeding on coralline algae. This was followed by a phase of rapid growth which led to sexual maturity at the end of the second year. The starfish were in the vicinity of 200 mm diameter at this stage. They continued to grow rapidly and were in the order of 300 mm at 3 years of age. Then they reached a plateau between 3 and 4 years and tended to decline after 4 years. Gonad development was greater in the third and subsequent years than at 2 years and there was a seasonal pattern of gametogenesis and spawning with water temperature being the only apparent cue in the indoor aquarium. Most specimens of A. planci died from ‘senility’ during the period 5-7.5 years, i.e. they fed poorly and shrank.

Field observations of life-cycle

The data above are derived from laboratory studies of A. planci, which are much more readily obtained than equivalent data from the field. The laboratory observations, however, accord with the limited field observations of life-cycle.

As in laboratory studies where A. planci larvae were found to select coralline algae for settlement, early juveniles (<20 mm diameter) were found on subtidal coralline algae (Porolithon onkodes) on the windward reef front of Suva Reef (Fiji). The juveniles were found in a variety of habitats where they were highly concealed: under coral blocks and rubble in the boulder zone of the exposed reef front; on dead bases of Acropora species in more sheltered areas; in narrow spaces within the reef crest; and on the fore-reef slope to depths of 8 m.

Growth rates on Suva Reef were found to be 2.6, 16.7 and 5.3 mm/month increase in diameter in the pre-coral feeding, early coral feeding and adult phases, respectively. This is in accord with the sigmoidal pattern of size versus age observed in laboratory studies, i.e. slow initial growth, a phase of very rapid growth beginning at coral feeding and tapering off of growth after the starfish reaches sexual maturity. In reefs in the Philippines, female and male specimens matured at 13 and 16 cm respectively.

Stump identified bands in the upper surface spines of A. planci and attributed these to annual growth bands. He didn't report growth rates based on these age determinations, and mark and recapture data, but he reported that the growth bands revealed 12+ year-old starfish: much older than those that became ‘senile’ and died in the laboratory.

In a small number of field studies, mortality rates of juvenile A. planci have to found to be very high, e.g. 6.5% per day for month-old and 0.45% per day for 7-month-old. Most of the mortality comes from predators, such as small crabs, that occur in and on the substrate with the juveniles. It is possible, however, that these rates may not reflect mortality over the range of habitats occupied by small juveniles.

Ecological impact on reefs

Popular anxiety to news of high densities of A. planci on the Great Barrier Reef was reflected in many newspaper reports and publications such as ‘Requiem for the Reef', which also suggested that there was a cover-up of the extent of damage. There was a popular idea that the coral and with it whole reefs were being destroyed by the starfish. In fact, as described in section 3.2 Behaviour, the starfish preys on coral by digesting the surface of living tissue from the coral skeletons. These skeletons persist, together with the mass of coralline algae that is essential for reef integrity. The initial change (first order effect) is loss of the veneer of living coral tissue.

A. planci is a component of the fauna of most coral reefs and the effects of A. planci populations on coral reefs are very dependent on the population density. At low densities (1 to perhaps 30/hectare) the rate at which coral is being preyed upon by the starfish, is less than the growth rate of the coral, i.e. the surface area of living coral is increasing. The starfish may, however, influence the coral community structure. Because the starfish don’t feed indiscriminately they may cause a distribution of coral species and colony sizes that differs from a pattern without them. This is evident by comparison of coral reefs where A. planci hasn’t been found to the more typical reefs with A. planci.

Some ecologists suggest that the starfish has an important and active role in maintaining coral reef biodiversity, driving ecological succession. Before overpopulation became a significant issue, crown-of-thorns prevented fast-growing coral from overpowering the slower growing coral varieties.

At high densities (‘outbreaks’, ‘plagues’), which may be defined as when the starfish are too abundant for the coral food supply, coral cover goes into decline. The starfish must broaden their diet from their preferred species, colony size and shape. The starfish often aggregate during feeding, even at low densities, but during high densities the cleared coral patches become almost continuous or completely continuous (photograph). There are second-order effects of these large areas of predated coral.

Aesthetically, in all the above cases, the reef surface is not as attractive as the living coral surface, but it is anything but dead.

There is a third-order effect potentially arising from the invasion by filamentous algae. Animals that depend directly or indirectly on hard corals, e.g. for shelter and food, should lose out, and herbivores and less specialist feeders gain. It would be expected that this would be most conspicuous in the fish fauna and long-terms studies of coral reef fish communities confirm this expectation. Observations of species abundance of coral reef fish communities considered 210 fish species from 10 reefs on the Great Barrier Reef. Over a period of 11 years, major disturbances, including outbreaks of A. planci and severe storms, resulted in major declines in coral cover on all reefs. Species abundance of coral reef fishes persisted on nine of the disturbed reefs. Extensive loss of coral resulted in declines in the group of species which were heavily coral-dependent, but this loss of diversity was more than compensated for by increases in the number of species that feed on invertebrates, algae and detritus. Because of this, species abundance in coral reef fish communities was estimated to be greatest at approximately 20% coral cover. Coral loss >20%, however, typically resulted in a decline in abundance of species in coral reef fish communities. The greatest impact, however, came from severe tropical storms, that resulted in an immediate loss of topographical complexity (photograph). Fish species from all trophic levels were more affected than by disturbances, such as A. planci predation, that kill corals but leave the reef framework intact.

Population outbreaks

Large populations of crown-of-thorns starfish (sometime emotively known as ‘plagues’) have been substantiated as occurring at twenty one locations of coral reefs during the 1960s to 1980s. These locations ranged from the Red Sea through the tropical Indo-Pacific region to French Polynesia. There were at least two substantiated repeated outbreaks at ten of these locations.

Values of starfish density from 140/ha to 1,000/ha have been considered in various reports to be outbreak populations, while starfish densities less than 100/ha have been considered to be low; however, at densities below 100/ha there may be feeding by A. planci that exceeds the growth of coral and there is net loss of coral.

From the surveys of many reef locations throughout the starfish's distribution large abundances of Acanthaster can be categorised as:

The Great Barrier Reef (GBR) is the most outstanding coral reef system in the world because of its great length, number of individual reefs and species diversity. When high densities of Acanthaster which were causing heavy mortality of coral were first seen about Green Island, off Cairns, in 1960–65, there was considerable alarm. High-density populations were subsequently found of a number of reefs to the south of Green Island, in the Central Great Barrier Reef region Some popular publications suggested that the whole Reef was in danger of dying: 'Requiem for the Reef' and 'Crown of Thorns: The Death of the Barrier Reef?'. They influenced and reflected some public alarm over the state and future of Great Barrier Reef.

There have been a number of studies modeling the population outbreaks on the GBR as a means to understand the phenomenon, e.g.

The Australian and Queensland governments funded research and set up advisory committees during the period of great anxiety about the nature of the starfish outbreaks on the GBR. They were regarded as not coming to terms with the unprecedented nature and magnitude of this problem and the two references above. Many scientists were criticised for not being able to give definitive but unsubstantiated answers. Others were more definitive in their answers Scientists were criticised for their reticence and for disagreeing on the nature and causes of the outbreaks on the GBR, hence the publication 'Starfish Wars' (cf. 'Star Wars').

Causes of population outbreaks

There was serious discussion and some strongly held views about the causes of this phenomenon. Some hypotheses focused on changes in the survival of juvenile and adult starfish - the "predator removal hypothesis":

Many of the reports of fish preying on Acanthaster are single observations or presumed predation from the nature of the fish. For example, the humphead wrasse may prey on the starfish amongst its more usual diet. Individual puffer fish and trigger-fish have been observed to feed crown-of-thorns starfish in the Red Sea, but there is no evidence that they are a significant factor in population control. A study, however, based on the stomach contents of large carnivorous fish that are potential predators of the starfish found no evidence of the starfish in the fish's guts. These carnivorous fish were caught commercially on the coral reefs on the Gulf of Oman and examined at local fish markets.

One problem with the concept of predators of large juvenile and adult starfish causing total mortality is that the starfish have good regenerative powers and they wouldn't keep still while being eaten. Also, they would need to be consumed completely or almost completely to die. 17–60% of starfish in various populations had missing or regenerating arms. Clearly the starfish experience various levels of sublethal predation. When the damage includes a major section of the disk together with arms, the number of arms regenerating on the disk may be less than the number lost.

Another hypothesis is the "aggregation hypothesis", whereby large aggregations of A. planci appear as apparent outbreaks because they have consumed all the adjacent coral. This seems to imply that there is apparently a dense population outbreak when there has already been a more diffuse population outbreak that has been dense enough to comprehensively prey on large areas of hard coral.

Female crown-of-thorns starfish are very fecund. Based on the eggs in ovaries, 200, 300 and 400 mm diameter females potentially spawn approximately 4, 30 and 50 million eggs, respectively (see also Gametes and embryos). Lucas adopted a different approach, focusing on the survival of the larvae arising from the eggs. The rationale for this approach was that small changes in the survival of larvae and developmental stages would result in very large changes in the adult population. Considering two hypothetical situations.

Twenty million eggs, from a female spawning and having a survival rate of about 0.00000001% throughout development would replace two adult starfish in a low-density population where the larvae recruit. If, however, the survival rate increases to 0.1% (one in a thousand) throughout development from one spawning of 20 million eggs this would result in 20,000 adult starfish where the larvae have recruited. Since the larvae are the most abundant stages of development it is likely that changes in survival will be most importance during this phase of development.

Temperature and salinity have little effect on the survival of crown-of-thorns larvae. However, abundance and species of the particular component of phytoplankton (unicellular flagellates) on which the larvae feed has a profound effect on survival and rate of growth. The abundance of phytoplankton cells is especially important. As autotrophs, phytoplankton abundance is strongly influenced by the concentration of inorganic nutrients, such as nitrogenous compounds.

Birkeland had observed a correlation between the abundance of crown-of-thorns on reefs adjacent to land masses. These occurred on mainland islands as distinct from coral atolls about three years after heavy rainfall that followed a period of drought. He suggested that runoff from such heavy rainfall may stimulate phytoplankton blooms of sufficient size to produce enough food for the larvae of A. planci through input of nutrients.

Combining Birkeland observations with the influence of inorganic nutrients on survival of the starfish larvae in experimental studies gave support for a mechanism for starfish outbreaks:

increased terrestrial runoff → increased nutrients denser phytoplankton↑→ better larval survival → increased starfish populations

There have been further conformations of these connections, however research by Olson (1987), Kaufmann (2002), and Byrne (2016) suggests terrestrial runoff has little or no impact on larval survival. The conflicting data describing the negligible role of terrestrial agricultural runoff has been described as "an inconvenient study".

There is also a flow-on effect in that where there are large starfish populations producing large numbers of larvae, there is likely to be heavy recruitment on reefs downstream to which the larvae are carried and then settle.

Population control

Population numbers for the crown-of-thorns have been increasing since the 1970s. However, historic records of distribution patterns and numbers are hard to come by, as SCUBA technology, necessary to conduct population censuses, had only been developed in the previous few decades.

To prevent overpopulation of crown-of-thorns causing widespread destruction to coral reef habitats, humans have implemented a variety of control measures. Manual removals have been successful, but are relatively labour-intensive. Injecting sodium bisulphate into the starfish is the most efficient measure in practice. Sodium bisulphate is deadly to crown-of-thorns, but it does not harm the surrounding reef and oceanic ecosystems. To control areas of high infestations, teams of divers have had kill rates of up to 120 per hour per diver. The practice of dismembering them was shown to have a kill rate of 12 per hour per diver and the diver performing this test was spiked three times. Therefore, it is for this reason and not rumours that they might be able to regenerate that dismembering is not recommended.

An even more labour-intensive route, but less risky to the diver, is to bury them under rocks or debris. This route is only suitable for areas with low infestation and if materials are available to perform the procedure without damaging corals.

A 2015 study by James Cook University showed that common household vinegar is also effective, as the acidity causes the starfish to disintegrate within days. Vinegar is also harmless to the environment, and is not restricted by regulations regarding animal products such as bile.

A new successful method of population control is by the injection of thiosulfate-citrate-bile salts-sucrose agar (TCBS). Only one injection is needed, leading to the organism's death in 24 hours from a contagious disease marked by "discoloured and necrotic skin, ulcerations, loss of body turgor, accumulation of colourless mucus on many spines especially at their tip, and loss of spines. Blisters on the dorsal integument broke through the skin surface and resulted in large, open sores that exposed the internal organs."

An autonomous starfish-killing robot called COTSBot has been developed and as of September 2015 was close to being ready for trials on the Great Barrier Reef. The COTSbot, which has a neural net-aided vision system, is designed to seek out crown-of-thorns starfish and give them a lethal injection of bile salts. After it eradicates the bulk of the starfish in a given area, human divers can move in and remove the survivors. Field trials of the robot have begun in Moreton Bay in Brisbane to refine its navigation system, according to Queensland University of Technology researcher Matthew Dunbabin. There are no crown-of-thorns starfish in Moreton Bay, but when the navigation has been refined, the robot will be used on the reef.